TY - JOUR
T1 - Levodopa-induced dyskinesia in Parkinson's disease
T2 - Insights from cross-cohort prognostic analysis using machine learning
AU - Loo, Rebecca Ting Jiin
AU - Tsurkalenko, Olena
AU - Klucken, Jochen
AU - Mangone, Graziella
AU - Khoury, Fouad
AU - Vidailhet, Marie
AU - Corvol, Jean Christophe
AU - Krüger, Rejko
AU - Glaab, Enrico
AU - Acharya, Geeta
AU - Aguayo, Gloria
AU - Alexandre, Myriam
AU - Batutu, Roxane
AU - Beaumont, Katy
AU - Berchem, Guy
AU - Boussaad, Ibrahim
AU - Contesotto, Gessica
AU - DE Bremaeker, Nancy
AU - Fritz, Joëlle
AU - Gantenbein, Manon
AU - Georges, Laura
AU - Graas, Jérôme
AU - Henry, Estelle
AU - Hundt, Alexander
AU - Jónsdóttir, Sonja
AU - Kofanova, Olga
AU - Lambert, Pauline
AU - Lorentz, Victoria
AU - Marques, Guilherme
AU - Mcintyre, Deborah
AU - Mediouni, Chouaib
AU - Menster, Myriam
AU - Mittelbronn, Michel
AU - Nickels, Sarah
AU - Noor, Fozia
AU - Pauly, Claire
AU - Pauly, Laure
AU - Pavelka, Lukas
AU - Perquin, Magali
AU - Pexaras, Achilleas
AU - Rauschenberger, Armin
AU - Roland, Olivia
AU - Sapienza, Stefano
AU - Sharify, Amir
AU - Sokolowska, Kate
AU - Theresine, Maud
AU - Thien, Hermann
AU - Trouet, Johanna
AU - Vaillant, Michel
AU - Vega, Carlos
AU - the NCER-PD Consortium
AU - Ferrari, Angelo
AU - Gamio, Carlos
AU - Henry, Margaux
AU - Festas Lopes, Ana
AU - Mendibide, Alexia
AU - Mtimet, Saïda
AU - Munsch, Maeva
AU - Remark, Lucie
AU - Richard, Ilsé
AU - Nehrbass, Ulf
N1 - Funding
EG acknowledges support by the Luxembourg National Research Fund (FNR) for the project RECAST (INTER/22/17104370/RECAST) as part of the Joint Programme - Neurodegenerative Disease Research (JPND) and for the project PreDYT (INTER/EJP RD 22/PREDYT) as part of the EJP RD Joint Transnational Call 2022 (JTC 2022). OT and JK acknowledge support by the FNR for the project “Digital Healthcare Solutions: Patient Management e-Health Concepts” (dHealthPD 14146272). The National Centre of Excellence in Research on Parkinson's Disease (NCER-PD) received funding from the Luxembourg National Research Fund (FNR/NCER13/BM/11264123). PPMI - a public-private partnership - is funded by the Michael J. Fox Foundation for Parkinson's Research and funding partners. A list of names of all of the PPMI funding partners can be found at sponsors/www.ppmi-info.org/about-ppmi/who-we-are/study-sponsors/ . The ICEBERG cohort received funding and support from the Agence Nationale de la Recherche (ANR) under grant agreements ANR-10-IAIHU-06 (IHU ICM), association France Parkinson, and the Fondation d’Entreprise EDF, and the Fondation Saint Michel, and Energipole.
Publisher Copyright:
© 2024 The Author(s)
PY - 2024/9
Y1 - 2024/9
N2 - Background: Prolonged levodopa treatment in Parkinson's disease (PD) often leads to motor complications, including levodopa-induced dyskinesia (LID). Despite continuous levodopa treatment, some patients do not develop LID symptoms, even in later stages of the disease. Objective: This study explores machine learning (ML) methods using baseline clinical characteristics to predict the development of LID in PD patients over four years, across multiple cohorts. Methods: Using interpretable ML approaches, we analyzed clinical data from three independent longitudinal PD cohorts (LuxPARK, n = 356; PPMI, n = 484; ICEBERG, n = 113) to develop cross-cohort prognostic models and identify potential predictors for the development of LID. We examined cohort-specific and shared predictive factors, assessing model performance and stability through cross-validation analyses. Results: Consistent cross-validation results for single and multiple cohort analyses highlighted the effectiveness of the ML models and identified baseline clinical characteristics with significant predictive value for the LID prognosis in PD. Predictors positively correlated with LID include axial symptoms, freezing of gait, and rigidity in the lower extremities. Conversely, the risk of developing LID was inversely associated with the occurrence of resting tremors, higher body weight, later onset of PD, and visuospatial abilities. Conclusions: This study presents interpretable ML models for dyskinesia prognosis with significant predictive power in cross-cohort analyses. The models may pave the way for proactive interventions against dyskinesia in PD by optimizing levodopa dosing regimens and adjunct treatments with dopamine agonists or MAO-B inhibitors, and by employing non-pharmacological interventions such as dietary adjustments affecting levodopa absorption for high-risk LID patients.
AB - Background: Prolonged levodopa treatment in Parkinson's disease (PD) often leads to motor complications, including levodopa-induced dyskinesia (LID). Despite continuous levodopa treatment, some patients do not develop LID symptoms, even in later stages of the disease. Objective: This study explores machine learning (ML) methods using baseline clinical characteristics to predict the development of LID in PD patients over four years, across multiple cohorts. Methods: Using interpretable ML approaches, we analyzed clinical data from three independent longitudinal PD cohorts (LuxPARK, n = 356; PPMI, n = 484; ICEBERG, n = 113) to develop cross-cohort prognostic models and identify potential predictors for the development of LID. We examined cohort-specific and shared predictive factors, assessing model performance and stability through cross-validation analyses. Results: Consistent cross-validation results for single and multiple cohort analyses highlighted the effectiveness of the ML models and identified baseline clinical characteristics with significant predictive value for the LID prognosis in PD. Predictors positively correlated with LID include axial symptoms, freezing of gait, and rigidity in the lower extremities. Conversely, the risk of developing LID was inversely associated with the occurrence of resting tremors, higher body weight, later onset of PD, and visuospatial abilities. Conclusions: This study presents interpretable ML models for dyskinesia prognosis with significant predictive power in cross-cohort analyses. The models may pave the way for proactive interventions against dyskinesia in PD by optimizing levodopa dosing regimens and adjunct treatments with dopamine agonists or MAO-B inhibitors, and by employing non-pharmacological interventions such as dietary adjustments affecting levodopa absorption for high-risk LID patients.
KW - Cross-cohort analysis
KW - Levodopa-induced dyskinesia
KW - Longitudinal cohorts
KW - Machine learning
KW - Predictive modeling
KW - Prognosis
UR - http://www.scopus.com/inward/record.url?scp=85198036170&partnerID=8YFLogxK
UR - https://pubmed.ncbi.nlm.nih.gov/38991633/
U2 - 10.1016/j.parkreldis.2024.107054
DO - 10.1016/j.parkreldis.2024.107054
M3 - Article
AN - SCOPUS:85198036170
SN - 1353-8020
VL - 126
JO - Parkinsonism and Related Disorders
JF - Parkinsonism and Related Disorders
M1 - 107054
ER -